Distance Sampling
Distance Sampling for Estimating Abundance · Also known as: line transect, point transect, distance estimation, detection probability
Distance sampling is a statistical method for estimating population abundance from data on distances between observers and detected individuals. Developed by Buckland and colleagues (1993) and formalized in the software Distance, this approach accounts for imperfect detection: animals far from an observer are less likely to be detected. By modeling the detection function (probability of detecting an animal at various distances), distance sampling produces unbiased estimates of abundance and density even when detection is incomplete.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use distance sampling to estimate wildlife abundance in line or point transect surveys, especially for mobile or cryptic species where visual detection varies with distance. Ideal for large study areas where complete surveys are infeasible. Requires accurate distance measurements from observer to object and reasonable assumptions about a smooth, monotonically declining detection function. Not appropriate if your study organism is sedentary or if distances cannot be measured reliably.
Strengths & limitations
- Produces less-biased abundance estimates than simple counts because it explicitly models the detection process
- Works with imperfect detection and accounts for animals missed during surveys
- Applicable to many sampling designs (line transects, point transects, cue counts) and taxa (birds, mammals, plants, insects)
- Provides confidence intervals and goodness-of-fit tests to assess model reliability
- Computationally efficient and implemented in accessible software (Distance, R packages Distance or unmarked)
- Accuracy depends critically on correctly specifying the detection function; poor fits lead to biased abundance estimates
- Assumes that distance measurements are exact, which is often difficult in the field (especially for cryptic species far away)
- Requires a sufficiently large sample of detections to fit the detection function reliably; sparse data reduce precision
- Does not handle complex spatial patterns well (e.g., clustering of animals) unless explicitly modeled
- Sensitive to violations of assumptions: objects at zero distance must be detected with certainty, and animals must not move in response to the observer
Frequently asked
How should I choose between line transects and point transects?
Line transects are more efficient for wide-ranging, mobile species in large open areas (savanna, open water). Point transects are better for stationary or slow-moving species and small study areas. Line transects require observers to walk steadily and record distances perpendicular to the path; point transects require stationary observation. Both produce unbiased estimates if assumptions are met. Pilot studies help determine which suits your target species and habitat.
What do I do if my distance data have a spike at zero distance?
A spike at zero distance (many animals detected very close) is common and usually acceptable. It does not violate the assumption of certainty of detection at zero. However, a spike far from zero (heaping at round numbers like 10, 25, or 50 meters) suggests measurement error or observer bias. You can use adjustment terms (polynomials) in the detection function or use distance classes instead of exact distances.
How many transects and detections do I need?
As a rule of thumb, aim for at least 60 to 100 detections to fit a stable detection function. Fewer detections lead to wide confidence intervals and unreliable estimates. The number of transects depends on detection rate; sparse populations require more transects. Power analysis tools in Distance software help determine required sampling effort for a target precision.
What happens if I violate the assumption that animals do not move in response to my presence?
If animals flee or move away from you as you approach, detected distances are biased upward, and abundance is underestimated. Conversely, if animals are attracted to you, distances are biased downward, and abundance is overestimated. Document observer-induced movement during pilot studies. If movement is severe, consider indirect methods (sign surveys) or adjust analysis to account for movement.
Can distance sampling handle spatial clustering of animals?
Distance sampling assumes that detections are independent. If animals are highly clustered (e.g., herds or flocks), you can record cluster distance and size, then model cluster density separately. More advanced spatial distance sampling methods (Density Surface Modelling) account for large-scale environmental covariates that explain density variation across the landscape.
Sources
- Buckland, S. T., Anderson, D. R., Burnham, K. P., Laake, J. L., Borchers, D. L., & Thomas, L. (1993). Distance Sampling: Estimating Abundance of Biological Populations. Chapman and Hall, London. link ↗
- Thomas, L., Buckland, S. T., Rexstad, E. A., et al. (2010). Distance software: design and analysis of distance sampling surveys for estimating population size. Journal of Applied Ecology, 47(1), 5-14. DOI: 10.1111/j.1365-2664.2009.01737.x ↗
- Buckland, S. T., Anderson, D. R., Burnham, K. P., Laake, J. L., Borchers, D. L., & Thomas, L. (2001). Introduction to Distance Sampling. Oxford University Press. DOI: 10.1093/oso/9780198506492.001.0001 ↗
How to cite this page
ScholarGate. (2026, June 3). Distance Sampling for Estimating Abundance. ScholarGate. https://scholargate.app/en/ecology/distance-sampling
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Food Web TopologyEcology↔ compare
- Leslie MatrixEcology↔ compare
- Population Viability AnalysisEcology↔ compare
- Species AccumulationEcology↔ compare